Clustering in Data Mining (DM) plays a substantial role in solving problems of data analysis in business and scientific applications. However, it has been a challenging factor in clustering Big-Data (BD) [5], as and then there is a rapid growth in size of datasets in scale of extra in the real world. The efficient way of solving the BD problem is to use a Map-Reduced with a desirable parallel data analysis with a widely used field of data processing. Hadoop provides an environment of cloud and usually used analysis utensil for big data. K-Means and Fuzzy K Means a parallelized Big Data Analysis (BDA) tool in cloud environment. The de-merits of K-Mean parallelized algorithm is very much data sensitive to noisy, sensitive to basic condition and also relate to certain fixed shape, whereas Fuzzy K-Mean clustering has a basic issue related to computing and processing time, and also is much complex than K-Means, our work presents an empirical study on present clustering methods which are used in BDA [4]. Empirical analyses of the present techniques are carried out with the appropriate methods of clustering and have been studied